May 24, 2024 - Science

AI uncovers how plant science evolved

an illustration of a robot hand holding a stem of green leaves

Illustration: Tiffany Herring/Axios

A new AI analysis of plant biology papers reveals what research topics countries are prioritizing and how different tools and technologies have steered the field.

Why it matters: AI is often touted as a tool to help scientists in a key aspect of their work: keeping up with a deluge of scientific papers.

  • There are tens of thousands of papers published on different topics, their findings described in an exacting and nuanced language that can require expert proficiency to decipher.
  • "Hopefully this will be an initial study to help people organize their own set of knowledge," study co-author Shin-Han Shiu of Michigan State University says.

What they found: Using machine learning and language algorithms to analyze nearly 422,000 abstracts of plant science papers, they found the evolution of the field of plant biology is often driven by technology that matures or reduces in cost.

  • That was seen in a few major transitions in the field in the early 1980s as techniques for cloning genes became more widely used and a decade later as the tools for developing mutant plants improved, they write in PLoS Biology today.
  • The most dramatic shift though was in the early 2000s, when genome sequencing became less costly.

Between the lines: The impact of AI, which is increasingly used in the field but only in the last 10 years, still isn't seen, Shiu says, adding he expects it see its effect on the field reflected in the literature in the next five years.

The big picture: The researchers also analyzed differences in research topics between countries.

  • For example, new plant compounds associated with herbal medicine are a focus in China but not the U.S.
  • They also found the U.S. is being overtaken by China as the top publisher in plant sciences, and there is a dramatic improvement in the quality of the papers from China measured by their impact in the field, Shiu says. This echoes broader trends about improved quality in papers from China documented in other research.
  • These trends reflect research priorities and investments and could inform science policy and the funding of certain topics, they write.
Go deeper